The link between classical reserving and granular reserving through double chain ladder and its extensions

Hiabu, Munir; Margraf, Carolin; Martínez-Miranda, María Dolores; Nielsen, Jens Perch

doi:10.1017/s1357321715000288

Cited by 15 publications

(11 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, recent research strongly promotes claims reserving on individual claims data. [1][2][3][4][5][6][7][8][9] All these contributions assume a fixed and parametric structural form for which the distribution hypotheses have to be discussed and tested (eg, the work of Pigeon et al 6 assumes a multivariate skew normal distribution to the claim payments). Such fixed structural forms are moreover not very flexible and are sometimes very difficult to estimate due to complex likelihood functions.…”

Section: Introductionmentioning

confidence: 99%

A machine learning approach for individual claims reserving in insurance

Baudry¹,

Robert²

2019

Appl Stoch Models Bus & Ind

View full text Add to dashboard Cite

Accurate loss reserves are an important item in the financial statement of an insurance company and are mostly evaluated by macrolevel models with aggregate data in run‐off triangles. In recent years, a new set of literature has considered individual claims data and proposed parametric reserving models based on claim history profiles. In this paper, we present a nonparametric and flexible approach for estimating outstanding liabilities using all the covariates associated to the policy, its policyholder, and all the information received by the insurance company on the individual claims since its reporting date. We develop a machine learning–based method and explain how to build specific subsets of data for the machine learning algorithms to be trained and assessed on. The choice for a nonparametric model leads to new issues since the target variables (claim occurrence and claim severity) are right‐censored most of the time. The performance of our approach is evaluated by comparing the predictive values of the reserve estimates with their true values on simulated data. We compare our individual approach with the most used aggregate data method, namely, chain ladder, with respect to the bias and the variance of the estimates. We also provide a short real case study based on a Dutch loan insurance portfolio.

show abstract

Section: Introductionmentioning

confidence: 99%

A machine learning approach for individual claims reserving in insurance

Baudry¹,

Robert²

2019

Appl Stoch Models Bus & Ind

View full text Add to dashboard Cite

show abstract

“…Finally, some researchers have done comparisons of 18 between individual and collective approaches, often attempting to answer the question "what is the best approach?" (see Hiabu et al (2016); Huang et al (2015) or Charpentier and Pigeon (2016) for some examples).…”

Section: Introductionmentioning

confidence: 99%

Individual Loss Reserving Using a Gradient Boosting-Based Approach

Duval

Pigeon

2019

Risks

View full text Add to dashboard Cite

In this paper, we propose models for non-life loss reserving combining traditionalapproaches such as Mack’s or generalized linear models and gradient boosting algorithm in anindividual framework. These claim-level models use information about each of the payments madefor each of the claims in the portfolio, as well as characteristics of the insured. We provide an examplebased on a detailed dataset from a property and casualty insurance company. We contrast sometraditional aggregate techniques, at the portfolio-level, with our individual-level approach and wediscuss some points related to practical applications.

show abstract

“…Two other methods of incorporating expert knowledge of incurred data into these full cash flow models have been introduced in Martínez-Miranda et al (2013b) and Hiabu et al (2016). The first of these two methods is extracting the inflation of the cost of a single claim from the incurred data and then incorporates that information in the double chain ladder model of Martínez-Miranda et al (2012).…”

Section: Introductionmentioning

confidence: 99%

“…This estimate is the best the claims department of an insurance company (with all the expert knowledge on the nature and severity of each claim available in such a department) is able to do. Hiabu et al (2016) therefore produced a version of double chain ladder reproducing exactly the expert judgement of the RBNS reserves.…”

Section: Introductionmentioning

confidence: 99%

Cash flow generalisations of non-life insurance expert systems estimating outstanding liabilities

Hiabu

Margraf

Martínez-Miranda

et al. 2016

Expert Systems with Applications

View full text Add to dashboard Cite

For as long as anyone remembers non-life insurance companies have used the so called chain ladder method to reserve for outstanding liabilities. When historical payments of claims are used as observations then chain ladder can be understood as estimating a multiplicative model. In most non-life insurance companies a mixture of paid data and expert knowledge, incurred data, is used as observations instead of just payments. This paper considers recent statistical cash flow models for asset-liability hedging, capital allocation and other management decision tools, and develops two new such methods incorporating available incurred data expert knowledge into the outstanding liability cash flow model. These two new methods unbundle the incurred data to aggregates of estimates of the future cash flow. By a redistribution to the right algorithm, the estimated future cash flow is incorporated in the overall estimation process and considered as data. A statistical validation technique is developed for these two new methods and they are compared to the other recent cash flow methods. The two methods show to have a very good performance on the real-life data set considered.

show abstract

The link between classical reserving and granular reserving through double chain ladder and its extensions

Cited by 15 publications

References 18 publications

A machine learning approach for individual claims reserving in insurance

A machine learning approach for individual claims reserving in insurance

Individual Loss Reserving Using a Gradient Boosting-Based Approach

Cash flow generalisations of non-life insurance expert systems estimating outstanding liabilities

Contact Info

Product

Resources

About